The overall goal of this project is to generate fine-structure RNA maps in human and mouse (C57BL/6NJ) tissues and primary cell lines using a variety of high-throughput sequencing platforms, to evaluate the biological importance of novel transcripts by determining if evidence of their translated products can be identified. From each sample analyzed, we propose to isolate long (>200 nucleotides) and short (< 200 nucleotides) RNA in biological duplicate. Illumina-based maps for these samples will initially be generated using (1) RNA sequencing (-seq) of ribosomal (r-)RNA depleted long total RNA. (2) RNA-seq of tobacco acid pyrophosphatase (TAP) pre-treated short RNA (3) Pair-end Cap Analysis of Gene Expression (PE-CAGE) of total RNA. Additionally, for a subset of primary cell lines we will generate the above libraries from nuclear and cytoplasmic subcellular fractions. Long RNA-Seq data will be distilled down into functional elements consisting of splice junctions, polyadenylatio sites and de novo genes and transcripts. The short RNA data will be distilled into contigs representing the 5' ends of short RNAs up to the read length. PE-CAGE data will be analyzed to form clusters representing the 5' ends of transcripts linked to a tag internal to the transcript body. Importantly, each element will be assessed for reproducibility using a nonparametric Irreproducible Detection Rate (nplDR) script. Collectively, these data will allow for the detection of novel transcribed regions and supportive information as to the location of promoter regions and subcellular residence of transcripts. In aggregate, these data will be used to generate models of both noncoding and protein coding transcripts and to distinguish isoforms at complex loci necessary to obtain a comprehensive view of mammalian transcriptomes. For a subset of these samples we will simultaneously collect the genome sequence of the human donors to provide a reference map that will be used to map the RNA data against and derive information concerning allele-specific expression and RNA editing. Unannotated transcript models will be tested using long-range (PacBio/454) sequencing. Lastly, proteogenomic analysis will be done and the results compared against the unannotated transcripts.